Our understanding of how individuals change substance use behaviors after being in jail, the mechanisms whereby this change occurs, and which interventions might work best for different individuals is minimal at best. As a result, relapse back to substance use is the norm with devastating effects on ex-inmates and their communities. The first step is to understand those factors that lead to short- and long-term changes in drug use after incarceration. To begin this process we will expand our pilot work conducting in-depth interviews with recently incarcerated individuals with substance use problems in order to delineate the cognitive and affective mechanisms of action that mediate between the event of incarceration and substance use behavior change. Results of these interviews will be used to develop scales to measure the different constructs of the proposed model. Scale items will then be refined using cognitive interviews. The final instrument will be administered to 400 inmates. The resulting data will be used in psychometric analyses to examine the dimensionality (internal validity) of the individual constructs, confirm and refine the resulting scales, calculate reliability coefficients for each scale, and to examine the relationship of the model constructs to each other. The US correctional system has experienced exponential growth in the past three decades with the vast majority of incarcerations related to alcohol or illicit substance use. Substance use among State prisoners is approximately 27 times greater than in the general population. Despite this ever increasing trend of incarcerating individuals with substance use disorders little is known about the cognitive and affective mechanisms by which incarceration may impact substance use behaviors. The Sentinel Events Model is a dynamic model predicated on the observation that an attempt to stop using a drug is often preceded by a negative consequence of use, such as the experience of a health, interpersonal, or legal problem (hitting bottom). We seek to better delineate the cognitive and affective mechanisms of action that mediate between a sentinel event (incarceration or a teachable moment) and behavior change. To test this model fully we need to first have validated measures of the model's constructs. These measures will assess several constructs, some of which may change rapidly over time. To capture changes over time we need measures than can be administered repeatedly during incarceration and after release. In this R21 we will create and validate measures of the model's constructs and examine the feasibility of different methods for obtaining ecological momentary assessment in jail and in the community. To this end, our specific aims are: SA 1: To understand and refine the model's constructs through in-depth interviews. SA 2: To create and validate measures of those constructs. SA 3: To refine procedures using these validated measures with EMA to determine the optimal frequency and depth of assessments.